THE CLOSING PLENARY COMMENCED ON THE 19TH OF NOVEMBER, 2010, AT 11:00 A.M. AS FOLLOWS:
ROB BLOKZIJL: Good morning. I would like to start this last session of RIPE 61, so if people in the back can make a decision, either to sit down or continue their conversations outside. Then, I think we can go.
There are a few small items for this last session. The first one, let me explain that you may have noticed that the RIPE NCC has launched a new project, the Atlas Project, measurements on a very large scale. A lot of you have picked up probes to bring home and connect. Planning these things is ?? has shown to be more complicated than we could ever dream, so we ran out of probes yesterday, or so.
That is the good news. Now, the bad news. Some of you will be requested to hand in your probes. Let me explain: A group of our colleagues from Samara in Russia have registered for probes but we have no more probes. Normally, we would say, it doesn't matter, we will ship them next week. However, shipping these mysterious little boxes full of electronics into Russia, is a whole other project. So Daniel came up with a brilliant solution, as usual, and that is as follows: If you live in Western Europe and you have a probe in your pocket, please be a volunteer and hand it back. I think Daniel needs eight probes. Hand them back. We promise that, next week, in the mail, you will find a real fresh probe and there is a competition in connection with ?? for the early participants and Daniel assures me that this will ?? you will not be disadvantaged in the competition for an iPad if you hand in your probe. Right now, I thought Daniel, he should say you get double entry in this lottery, but that is not my project. So, the request is: If you live in Western Europe, for instance like in the Netherlands, that makes shipping very rapid, or in Germany, almost as quick, and you have a probe and you want to help your colleagues in Samara to be on this network when they fly home today or tomorrow, please you have three or you need three? Another five. Four. OK. Five, you have thank you, Wilifred. All right. I think the message is understood and the last two or three probes, I'm sure Daniel will get.
Thank you all. This is what I call good cooperation.
Next on my little list. At the last ?? at every RIPE meeting, it's tradition that on Thursday at lunchtime, the Working Group chairs have lunch together and they discuss things like what went wrong at this meeting, logistics and everything. At the last RIPE meeting, our focus shifted a little bit and we thought let us do an evaluation of RIPE meetings, how the meetings are organised, whether it's still ?? still attracts the people we want to attract and the way forward is not to sit with, I think, about 30 Chairs and co?Chairs around the table and have endless discussions. So we found a few volunteers amongst ourselves who are a task force and their task is to do the thinking and come with suggestions.
So next on the agenda, I would like to call Andy who heads this task force and who will explain what ?? about the work that has been done and next steps.
ANDY DAVIDSON: Thank you for the opportunity to explain some of the things that we have been working on in the task force. I am going to feedback on two items rather than just the task force but I promise to do it quickly as I can.
First of all, I'd like to call about the RIPE BoF that we held yesterday for the first time after the final Working Group sessions and then I will go and talk about the activities of the task force, whether I think we have been a success to date and what we are going to do next.
So, yesterday, at 5:45 we held a RIPE BoF on the subject of IPv6 access, and the thing that makes the BoF format interesting is that it, unlike much of the agenda of the week, there is no set list of speakers, there are no slides, there are no bullet?points; there is simply people with problems and solutions, on different people with solutions to those problems and ideas and information that is going to drive the community's knowledge or motivation forward. So unlike a ?? unlike a session where someone is standing at the front and talking to a room of people, this is where the room talks to each other. So the style is very informal and very conversational and I learned ?? I work in the ?? in operations in the areas of v6 access is one that is interesting to me and I thought I understood the challenges, but I learned there were a lot more challenges that I'd understood on the way into the meeting, but, at the same time, I learned that a lot of the things I was worried about are actually solved and how to go ahead and work on something else now.
So I thought it was a really good meeting and maybe ?? were there the people in the BofF are they in the room now as well? I can see some hands. Keep your hand up if you think we should do that again on maybe a different topic? There are more hands now, so that is brilliant. Thank you very much. So that was the ?? that was how the RIPE BoF worked and what we did there.
The next thing are the activities of the task force. These are the things ?? there is two sides on this slide that were new to this RIPE meeting. First of all, the tutorial from someone in the community, that was the subject on ?? that was a tutorial on IPv6 security and this was another radical success, I believe. There was 100 or so people in this room watching a tutorial that came from someone in the community, delivered to the community, and we did a survey after that and the message was do more things of this type at RIPE meetings. So we are going to take that as a message and find more interesting topics to deliver tutorials on.
The BoF was the other activity that we worked on. But the one thing we are committed to do that we have started work on now, is survey the community, which is the people in this room, the people who have been to RIPE meetings before and the people we think should be at RIPE meetings because there are colleagues or there are friends of industry and they have good opinions and they will be welcome in this room and we need to find out what everybody wants from the RIPE meeting and that is the things that we care about are what must not change, what are the things that make this week in this format a success and we preserve those and keep those the same, but what would you like that is new? What we understand what the priorities that people want that we don't do now, the people I am working with, we are going to table that to the Working Group Chairs and the community at large to work out how we can put in place the things that you would like to see that we don't do now.
So it's very important to stress that this is very, very open, everybody is going to be able to give their opinion and everybody is able to shape the output of this. So this is your meeting and it's ?? and we are working for you to give you the meeting that you want.
So this is the process. The survey will be finished and released very soon. We are planning to feedback in January. And after that, we don't know. That depends on the output of the survey, but I am fairly sure it's going to be fun working on it.
I have tried to keep this brief. Are there any questions from the room? OK. Well, thank you.
ROB BLOKZIJL: I have a question. Do you have a point of contact where people on their way home this afternoon, say, oh, blimey, I should have told this or asked that. What is the ?? the e?mail address?
ANDY DAVIDSON: Given Brian suffered everyone in the task force to do this on Monday. I'm going to repeat it. Jim and Brian, if you are in the room and, if you can stand up you can talk to these people on the way, and there is ?? is everyone else from the task force in the room? Maybe Christian ?? well, you can talk to us because the other people in the task force may have had to leave already but you can also e?mail me or came Camilla or Rob or anyone who can forward the message. This is open; it's not a closed party. This is an open party. And although there will be some structure to asking the questions, there is an opportunity to just free form write to somebody in the community, maybe Rob or Camilla or myself, my e?mail address is up there and I will forward that to the whole group and we will discuss every opinion that you have.
And I look forward to hearing from you so please do do that, grab us in person or drop somebody an e?mail. OK.
ROB BLOKZIJL: Thank you.
Next on my little agenda is a report from Erik. You have all been ?? you may have noticed we had networking here and some of you are still connected, I see. Erik will explain what went on behind the scenes and a little bit about how much use we made of all that.
ERIK ROMIJN: OK. Good morning. I am Erik, this is the RIPE 61 technical report. I will start with a short introduction. We have eight engineers working on site here and a manager and someone supporting us in the office. A new team member this time is Bernadetto. Our task involved anything that has wiring, except beamers lighting and Stenography. Webcasts presentations systems and of course the network is all our work, registration desk plasma screens. This time, we no longer have an audio cast, we replaced that by also having a webcast in the room downstairs. We also have a few things which do not have wires. But they are there. Our network set?up has two routers, we this time have two up links one from GARR and one from CASPUR. Fibre provided by /TKPWO*T and inside the ?? three networks for separate purposes, the public one you are all using now, the one that my is using and the private one for management and registration.
Up link this time was gigabit from GARR, the primary one and backup CASPUR of 100 megabits.
This is the panel downstairs. The hotel, we couldn't convince the hotel to remove the door so it didn't end up very clean. We started using nice modular cases for network equipment. This ?? each of them has a switch and router. Once we are done, reremove the cables, put the covers on and back to Amsterdam.
Most of the patch cables you see there are for access points.
We introduced a number of new setups this time. The presentation system we were using in the main room has now also started use in the side room, which allows us to switch presentations much smoother and easier. It did require some initial tweaking before everything worked out. Apparently, some of you make PowerPoint presentations that do not run smoothly only if you have a gigabite of RAM.
We also introduced remote switchers, so I no longer have to switch remotes, and basically it's a little USB switch that allows us to switch between the two presentation Macs.
We introduced new servers, these are the old ones. They are very loud. We sit working next to them and we we were considering to buy noise cancelling headphones and we had to ask the hotel to make sure there was very strong air co in the room due to the heat, they are very large. So we replaced them by this. The two on the right are two virtual machine servers that we use, one standby and one active, next is TTM box and to the left of that is the UPS we now have for the servers in case of power failures. We introduced iPad and iPhone stream this time. Basically getting a new ?? which was supported by Apple. We added video streaming for the break?out room. At this point very simple but it works. And we set up MAC?minis in the terminal room. Our old set?up was to rent PCs from some local place, but it took us a long of time. On one occasion we got them delivered with illegal version of Windows and viruses. So this is much more efficient. We also introduced dual?screen set up so people can test the presentations because these Macs have the same software as our laptops and presentation MAC?minis. So if it works there, it works here.
There were a few issues during the meeting. We ran into a particularly tricky problem with duplicate address detection. It is, however, not very visible. No one noticed until Thursday and we think it was already there at the last meeting and nobody noticed. Basically, in a duplicate address detection the IPv6 address is kept in tentative state until it knows there is no duplicates. That should happen in a few seconds. But in our case up to a minute and you have no running v6 connectivity at that time so if your laptop breaks you might not have v6 for a minute. It's a tricky problem we are so far thinking it is related by dropping of IPv6 multi?task buyer switches in certain conditions when sending data from to certain types of untaxed reports and we will continue to work on this. And thanks to Lorenzo and Jen from Google for noticing and helping us.
We had some issues in upstreams of our upstreams. There was some packet loss that we noticed in TTM on Wednesday morning to several destinations and we had issues reaching heise over v6. We have escalated that to our upstreams but this did not get entirely resolved in it time, and also beyond our control. We had a printer advising IPv6 but not allowing to you print on IPv6 so your laptop would wait for several minutes for the IPv6 time?out before actually trying over IPv4. For now, we resolved that by removing IPv6, but that is not the permanent fix.
We also had some some slowness on the RIPE 61 website in the first couple of days. On Tuesday we resolved that by extra memory and additional layers of caching for the web server which massively reduced the loads. A strange problem we have been having here is some of our access points, at least in the first few days, became unreachable so they are still running and accepting associations but will discard all your traffic and you can see this one, this is a graph of one of them, this happened at 6:00 in the morning so it's not like there were a lot of people here or a lot of associations. We found some improvement after we stopped doing power over ethernet on some of them so this is basically dodgy hotel cabling but not dodgy enough to not work at all. Eventually we did get everything stable and it has stayed stable since then.
As usual, we have deployed a TTM box in the hotel. It measures to about 100 boxes. This was our initial deployment on the balcony out of the window but there was another balcony overhead so we had to extend the length a bit so that it could reach. The building you see in the background is actually the United States embassy but they did not seem to have a problem with the set?up.
So, Alex does TTM stats always and I inspected a particular unusual statistic here. This is the incoming delay from the TTM in the Vatican to here and it's half a millisecond on IPv6 and over 20 on IPv4, which is the largest difference I have ever seen in measurements.
What we can actually see is that going from the meeting network to the Vatican it seems normal, it started at third HOP the latency delay, the incoming delay that is only affected. The outgoing delay is good. So looking from the Vatican box, on IPv6 we see it looks pretty normal, it's short and pretty low latency. Looking at IPv4, we see that it is actually routed to their transits and then goes to Milan and then comes back here again. So basically, IPv6 brings us five hops closer to heaven.
And TTM got a little brother this time, we deployed right Atlas probe in the meeting network as well. Our routers have a USB ports, I spent more time taking the picture than deploying the probe. It was ?? and RIPE Atlas was still at an early stage so we didn't see a lot. I did notice this one which was probably F?Root doing a node switch, very briefly, so brief the DNS switch didn't see it. Normal latency is very low but suddenly jumps up to over 20 so it was most likely a node switch.
This is the amount of viewers on the webcast, the blue bit is people from the Internet watching on IPv4, red is.IPv6 and the other colours from the meeting venue on v4 and 6 and from the RIPE NCC office on v4 and v6, so we can see that there is about, on average, 42 percent of the webcast users Ewing IPv6, the peak was ?? the stats do not include GM you can notice there is more people watching from the venue, the brown bit, during the parallel session, so also practically everybody using it from the venue was using IPv6.
On the up link, we have hit a peak of 50 megabits in 21 out, the average being eight and six. Someone started doing something here, overnight, somebody uploaded 50 gigabytes from the meeting network and continued to do quite a bit of traffic. So we are not exactly sure what happened there. It's also another problem, it is just interesting. For IPv6, we have managed to hit a peak of 10 Meg bites in and one?and?a?half out and comparing the averages we can say that this time 10.8 of our up link traffic is actually IPv6.
Last time that was 6.8 percent, so we have increased about 70 percent.
If we would actually continue that growth, by RIPE 66 we will no longer have IPv4 traffic on the network if we continue linearly it will continue until RIPE 81 before we no longer need IPv4.
For the wireless we this time deployed 24 base stations and that is 30 percent higher than the last meeting. We deployed them in higher density now and the venue also causes us to deploy more. We hit a peak of 340 associations which is actually what we had last meeting so it didn't get any busier. If we continue this trend of 30 percent more axes points every meeting by RIPE 72 we should have more than attendees. And one of the ways we achieve this high?density distribution, there are actually eight access points in this room but you can see only four unless you are very small.
So that is it. Are there any questions? (Applause)
ROB BLOKZIJL: Are there any questions? No. Then, on behalf of all of us, it's a very hearty thank you very much for the whole team, Erik. Great.
Right. We saw that all that traffic mainly originated from the attendees, so let's have a look at a few other statistics. Attendees: 430. That is a record. This does not count Erik and your crew; these are people who had time to go to meetings instead of crawling under chairs and broom cabinets. 430. That is three more than last meeting. So I should have made some of type of charts and see where that leads us in the future.
The number of newcomers, that is the next slide, I think. 145. And I think that is an interesting sample for Andy to do his survey on, because these are people who came fresh and we hope that newcomers remain coming, not as newcomers, of course, the next time, but that is something we are very, very interested in learning, people who come for the first time to a RIPE meeting, whether they find it attractive to keep coming.
We came from, as usual, from a whole host of countries and you see the distribution here. The largest bunch came from I think ?? there are too many blues here. Many people from many countries and that is the whole ?? that is the whole idea.
We are international, but we knew that already.
And we come from various organisations and our newcomers have seen similar slides where I explain that when we started with RIPE, it was almost 100 percent academic, research, educational world, because that is where the Internet started in Europe, and now we see that it's about 50/50, commercial ISPs and, still, a healthy distribution of important other organisations, important in the sense that they think that it is important to be present when we do the work which we do. And that is ?? has remained more or less constant over the last couple of meetings.
So, I think that is all the slides I have.
Erik mentioned it already, we had two web streams, NaMeX and ?? sorry, CASPUR and GARR. We had local hosts, and that is extremely important because without the help of the local hosts, we are not to organise a meeting that turns out to be so well?organised. So I think we should give a big round of applause for these two organisations and especially I want to mention two people, Maurizio Goretti and Francesco Ferrari who have worked incredibly hard, not only this week, they started a bit earlier before you came. So, thank you.
Next to the local hosts, we had sponsors who helped us to make this meeting as smooth as I think it was. Sponsors sponsor various things; COLT sponsored the local loops, GARR sponsored various things, and altogether, we have enjoyed a lot of nice socials this week and there was a lot of son certificate effort that went into that, as well. So thank you, all the sponsors.
While we are on the issue of thanks, before people start slowly drifting off to airports, I think I would like to say a few words about the RIPE NCC staff that organises these meetings, the people at the registration desk, the technical team, various other of our colleagues from the RIPE NCC, presentations they gave, I think we should give a very warm round of applause for our good friends in the RIPE NCC. Thank you.
The people who do the registrations they, for obvious reasons, they like early registrations; the earlier, the better, because it gives them more time, in case of problems, to fix problems.
For many RIPE meetings already we run this programme to encourage you to do your early registrations and by having a few prizes for the first three who register. Now, I know who the three are but where are the prizes? Camilla is sitting on the prizes, these Chairs are really multi?functional.
So, the first one on the list, that is me, but that is a test case. That is hard?wired in the software.
So the first registration was by Sascha Pollok from Germany. Is he here? I have no idea what it is but be careful.
And the next one on my list is Torben von Nolcken Germany.
And the third one is Sergi from the Czech Republic. And Camilla tells me your prizes are waiting for you at the registration desk.
The RIPE NCC had member service centre operational this week and people who dropped in with questions, issues they wanted sorted out, they left the office with a ticket with a number, and we have a prize draw on that. In an open and transparent way, close your eyes, number 3? No. Yes. Show your lottery ticket. Show your certificate. Franchesco from Telecom Italia.
Secret Working Group. As the time already indicates, don't ask me; it's secret.
NIGEL TITLEY: No need to transcribe this, this is secret.
ROB BLOKZIJL: Right. As usual this is the only report where we don't invite persons and comments.
I think this, as far as I am concerned, brings us to the end of this meeting. There is one little at this time bit I want to share with you: We have this NCC services centre and they broke a record this week between in creating a new LIR, an organisation walked in and said "I want to become an LIR" 45 minutes they walked out being an LIR and received their first allocation. I think we should keep this in time because sometimes people claim that RIPE NCC has become a bureaucracy and it takes ages to process things. That proves if parties understand what they are doing it is a very short procedure and I would like to take this opportunity to say congratulations.
AUDIENCE SPEAKER: Did they only give ?? have one application?
ROB BLOKZIJL: Details, details, details. We are open and transparent. You can find that all in the information services. And OK. Without further ado, I thank you all for coming and working very hard this week and I personally thought this was one of the more active RIPE meeting we had. I wish you a pleasant stay in Rome if you are staying for a couple of days and for all of you, have a safe journey home and see you all in Amsterdam. Thank you.
RIPE 61 then concluded.
LIVE CAPTIONING BY AOIFE DOWNES RPR
DOYLE COURT REPORTERS LTD, DUBLIN IRELAND.